- Architecture Insights
- Posts
- #013: ChatGPT Prompting (Framework Prompts For Design Projects)
#013: ChatGPT Prompting (Framework Prompts For Design Projects)
Welcome to this week’s edition of Architecture Insights. A weekly newsletter on artificial intelligence for architects, landscape architects, and designers.
Understanding the correct way to prompt and what to ask ChatGPT is crucial for getting the results you want. However, if you are simply seeking inspiration or rough guides, using framework prompting is a great way to achieve that.
This Past Week in AI
1. Midjourney releases Alpha. This is the web version of Midjourney, at the moment it is only available to “power users” if you’ve generated more than 10,000 images. It may be another few weeks until we see it available for all users.
Prompting
When prompting ChatGPT or an alternative large language model (LLM), it is crucial to keep in mind two important factors to ensure you achieve the best results:
Prompt Context: Initiate by introducing or providing context for your inquiry or instruction.
This will help ChatGPT understand the topic and provide relevant insights. It can be any information regarding design, site, and anything project related.
Constraints or Specifics: If necessary, add supplementary details or constraints to guide ChatGPT's responses. This allows for project-specific variables to be considered, resulting in more tailored and contextually relevant suggestions.
This is sometimes best used after you have prompted GPT a few times and get an idea of the type of answers you are receiving. If you don’t like a certain part of the response then re-prompt it and tell it not to answer with (x) in the response.
Framework Prompts:
What are framework prompts?
Framework prompts are prompts you use where the end goal of the answer is to act as a template to customize and fill in personalized information that GPT does not have access to.
Customize each prompt as needed. Take the answers from each prompt and use them as a building block for your final desired outcome.
Warm/Cold Outreach Email: “I'm an architectural designer and I want to write an email to new potential clients who inquire about my services. I want this to be friendly but also encourage them to book a kick-off consultation meeting with me.”
Project Schedule: “Could you provide an 18-month project schedule that will follow the following deadlines: 4 weeks - concept design/development, 4 weeks - 50% construction set submission, 2 weeks - 75%-90% construction set submission, 2 weeks - Tender submission, 6 weeks - construction.”
Community Engagement Questions: “Can you give me a list of 50 questions to ask community members during the community involvement stage in the design of a new community town center. The land is currently occupied by a large portion of settlers who have voiced concerns in the past about improper use of the land.”
Design Approach: "A city is planning to construct a new city hall. The emphasis is on creating a design that is welcoming and accessible to all members of the community. How would you approach the design to achieve both a functional space for government officials and an inviting environment for the community.”
Resources & Tools
3 free courses to learn AI:
Google - “Introduction to generative AI Learning”.
Microsoft - “Transform Your Business With AI”.
LinkedIn - “Career Essentials In Generative AI”.
AI Image of the week
![](https://media.beehiiv.com/cdn-cgi/image/fit=scale-down,format=auto,onerror=redirect,quality=80/uploads/asset/file/333c3798-8d73-451d-abd3-95cda1dda768/sampalmer_a_floating_home_off_the_beach_of_Greece_ancient_roman_51f0c6df-8c81-41c3-9a38-842de69f82b9.png?t=1702626500)
Thank you for reading this week’s issue, check past issues here. Share this newsletter with colleagues, friends, or anyone interested in the combined world of architecture and artificial intelligence.
Feedback is always appreciated, let us know how we did by responding to this email.
Until next Friday,
A.I.